IRC log for #koha, 2006-10-13

All times shown according to UTC.

Time S Nick Message
03:23 osmoze hello
04:30 hdl tumer around ?
05:15 osmoze who have configured a printer under koha ?
05:15 hdl HLT ;)
05:27 tumer hdl:looking for me?
06:32 chris hmm missed osmoze ... hlt, kapiti and rangitikei all have printers set up in koha, hlt is all unix, rangitikei is windows machines using samba and cups, and kapiti is a mix of windows and linux
06:39 chris
08:15 hdl kados around ?
08:17 kados hdl: briefly ... about to head to the gym
08:17 hdl yes : Got a problem with rel3 Zconn.
08:17 kados k ... what's the prob?
08:18 hdl Because it doesnot use parameter to get the right connection.
08:18 hdl whatever parameter you ask for afeter 1st connection.
08:18 kados here's what I have:
08:18 hdl It sends you back current connection
08:18 kados sub Zconn {
08:18 kados    my $self=shift;
08:18 kados    my $server=shift;
08:18 kados    my $async=shift;
08:18 kados    my $auth=shift;
08:18 kados    my $piggyback=shift;    my $syntax=shift;
08:19 kados second one is the 'server'
08:19 hdl Yes it is a parameter.
08:19 kados so "biblioserver" for biblios and "authorityserver" for authorities
08:20 hdl But when you search for existing Zconn and find one, you return the 1st connexion whatever it is...
08:20 hdl :D
08:20 kados ahh
08:20 kados right ... because the $context-> global object
08:20 hdl So you cannot connect to authorities then to biblios....
08:20 kados hmmm
08:20 hdl I suggest
08:20 hdl Zconn could be a ref to a hash.
08:21 hdl and when asking $server
08:21 hdl returns the connection of the right server.
08:22 hdl Zconn could INTERNALLY be a ref to a hash.
08:22 kados but that would mean it would connect to every server every time
08:22 kados what if I have 10 servers defined?
08:22 kados (the Zconn is also used for federated searches remember)
08:22 hdl No : it would store connexions into a 10 entries hash.
08:24 hdl In Context vars, Zconn would be a ref
08:25 hdl Zconn->{biblioserver} would contain connection to "biblios (whatever its name)" database
08:25 hdl ....
08:25 hdl Is it clear ?
08:26 kados sec ... phone call
08:26 hdl Sorry.
08:29 hdl kados : It perfectly works.
08:29 hdl I commit my solution.
08:29 hdl You will be able to see.
08:30 kados ok ... thx
08:30 hdl The problem would be for multpile connexions to the same zebra server.
08:30 hdl can this happen ?
10:23 kados hey cm, how goes it?
10:24 kados cm: just so you know, NPL has seen a drastic increase in speed at the circ desk ... transactions running under a second
10:24 cm hey there.  it goes.  :)  I have our opac up at  It should look familiar.  ;)
10:24 cm very cool about the transaction speed.
10:24 kados nice
10:25 kados cm: looks like your MARC records are in pretty bad shape
10:25 kados cm: is that what they looked like out of your current system?
10:26 cm mmm, yes.  what is so bad about them?
10:26 kados you're missing leaders and fixed fields for one
10:26 cm ah, wonderful.
10:26 kados those are _very_ important in Koha
10:26 kados all of the subtype limits rely on proper values ...
10:27 kados as do some of the sort by options
10:28 kados cm: do you have a data dump from your old system handy?
10:28 kados (before Kyle runs it through his merge routine?)
10:28 cm sure, I can dig one up.  I can post it on our web server so you can download it easily.
10:29 kados cool, let me know where
10:29 cm ok, give me a few.
10:29 cm minutes, that is.  :)
10:29 kados :-)
10:35 cm okay, here:
10:36 cm there are two.  mplcurrent.mrc is before I run it through marc2xml & xml2marc.  mplcurrentout.mrc is after.
10:37 kados cm: just so we're clear ... mplcurrent.mrc is what you get directly from your old ILS ... before Kyle imports into Koha 2.2 and merges the records?
10:39 kados cm: those records have leaders, and 008 fields ...
10:39 kados cm: so maybe your framework is the problem
10:40 kados cm: do you have framework entries for 000 (leader) and 008?
10:41 kados cm: also ... from your old system, were your itemtypes at the record level or item level?
10:41 kados I see a 961 field
10:42 kados 961    _tJBOC
10:42 cm it's what I get after tweaking them with marcedit, but I don't touch 000 or 008.  I'll check my framework.  i'll check record.abs too.
10:42 cm 961t is our item type.
10:42 kados it wouldn't be in record.abs
10:42 kados that's just for the index
10:43 kados ahh ... you tweak them in marcedit?
10:43 kados (btw: unless I'm mistaken, marcedit can natively export to UTF-8)
10:43 kados what do they look like befre marcedit gets ahold of them?
10:44 cm yes, but i had problems with marcedit's utf8 exporter.  iirc, it was crashy.
10:44 kados ahh ... actually, that makes me feel better :-)
10:44 cm I'll see if I can find a pre-marcedit copy and upload it.
10:44 kados our marc8 export routines crash sometimes too :-)
10:44 cm :)
10:45 cm 000 & 008 seem to be missing from my framework.  that would explain it!
10:45 kados yep :-)
10:45 kados you'll want to add those
10:45 kados in fact ... i suspect your framework is missing quite a lot
10:45 kados what I'd do is grab the latest marc21 framework from rel_2_2
10:45 kados it's a sql file
10:46 kados and just edit that sql file with the changes
10:46 kados that way you have a back up of your framework that you can rely on
10:46 kados (and besides, I have heard reports that the framework editor is buggy ... we don't use it at LibLime)
10:46 cm ok.  
10:47 kados by the way ... kudos on getting the system up and running ... no easy feat!
10:47 cm yeah, no kidding!
10:47 kados :-)
10:49 kados ok ... I'll be back soon
10:59 alaurin kados : i've got a problem with zebra indexation ....,
10:59 alaurin i've got this message

| Channels | #koha index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary